Chronic low back pain is a major public health concern. Psychosocial pain interventions have proliferated, and some approaches have strong support for efficacy. Most research has been focused on questions regarding the overall efficacy of these treatments. However, important questions regarding the mechanisms by which these psychosocial pain treatments produce benefits have been neglected. We have not yet empirically determined how psychosocial pain treatments work. Addressing this knowledge gap requires that we shift attention in psychosocial pain research away from evaluating only treatment efficacy and toward research that uncovers effective core treatment mechanisms. We advance and propose to test 2 conceptual models of treatment mechanisms. First, the Specific Mechanisms Model holds that different treatments work primarily via mechanisms specific to the treatment. Thus, Behavioral Treatment (BT) theoretically works in persons with pain via changes in behavioral coping, Cognitive Therapy (CT) works via changes in cognitive content, and MBSR works via changes in cognitive process. In contrast, a second approach holds that psychosocial pain interventions are effective because they all operate via 1 or more shared mechanisms, and can be termed the Shared Mechanisms Model. This would include mechanisms heralded by 1 or more of the specific theories and general mechanisms that are reputed to be factors held in common by all viable psychosocial interventions (e.g., therapeutic relationship). To test the validity of the 2 models and identify the mechanisms critical for the efficacy of psychosocial pain treatments, 400 people with chronic low back pain will be randomly assigned to 1 of 4 treatments: BT, CT, MBSR, or a Pain Education control condition. Eight sessions of 1.5 hrs duration will be delivered. Mechanism and outcome measures will be assessed weekly. We will first determine which model - specific vs. shared - best explains the data. In the event that the Shared Mechanism Model is supported, further analyses will determine which mechanism(s) is (are) critical across the treatments. Next, we will evaluate effects of mechanisms beyond pre-post changes and focus on 6-12 mos follow-up. It may be that the mechanisms showing effects on salutary pre-post outcomes will be the same mechanisms exerting effects on 6-12 mos maintenance and/or enhanced outcomes. It may also be that mechanisms not showing strong associations with pre-post outcome changes may emerge as important predictors of 6-12 mos outcomes. Results will shed important new light on what makes psychosocial pain treatments work. ous project or program of research has to date. Such findings will increase understanding of the causal impacts of each of the hypothesized mechanisms on outcomes, and so reveal which one or more mechanisms should be the primary target(s) of treatment procedures and techniques. This will allow us to determine whether in further interventions we should focus all or most of our efforts on changing what people with chronic pain do, the content of their thoughts, how they respond to their thoughts, just two of these, or all three.